ORIGINAL RESEARCH
published: 10 May 2019
Edited by:
Hasan Ayaz,
Drexel University, United States
Reviewed by:
Meltem Izzetoglu,
Villanova University, United States
Murat Ozgoren,
Dokuz Eylül University, Turkey
*Correspondence:
Wenfeng Wu
wwfhsd@aliyun.com
Received: 18 August 2018
Accepted: 23 April 2019
Published: 10 May 2019
Citation:
Lu Y, Wu W, Mei G, Zhao S,
Zhou H, Li D and Pan D (2019)
Surface Acting or Deep Acting, Who
Need More Effortful? A Study on
Emotional Labor Using Functional
Near-Infrared Spectroscopy.
Front. Hum. Neurosci. 13:151.
doi: 10.3389/fnhum.2019.00151
Frontiers in Human Neuroscience | www.frontiersin.org
doi: 10.3389/fnhum.2019.00151
Surface Acting or Deep Acting, Who
Need More Effortful? A Study on
Emotional Labor Using Functional
Near-Infrared Spectroscopy
Yongbiao Lu 1 , Wenfeng Wu 1* , Gaoxing Mei 1 , Shouying Zhao 1 , Haibo Zhou 2 , Daling Li 1
and Deng Pan 1
1
School of Psychology, Guizhou Normal University, Guiyang, China, 2 School of Education, Hunan University of Science
and Technology, Xiangtan, China
Emotional labor is characterized by two main regulation strategies: surface acting and
deep acting. However, which strategy consumes more energy? To explore this, we used
functional near-infrared spectroscopy (fNIRS) to measure changes in hemoglobin density
while participants performed a task requiring them to make the opposite emotional facial
expression of that presented in a picture. We found that (1) neither surface nor deep
acting led to a significant change in hemoglobin concentration in the prefrontal cortex;
(2) making negative and positive facial expressions activated the same left front and
middle areas of the prefrontal cortex; and (3) making positive facial expressions activated
the rear portion of the prefrontal cortex, but making negative facial expressions did not.
Based on these findings and past work, we can infer that deep and surface acting may
not significantly differ in terms of the activity in the prefrontal cortex energy consumed.
Furthermore, engaging in positive and negative emotional labor appear to utilize some
of the same neurological mechanisms, although they differ in others.
Keywords: emotional labor, surface acting, deep acting, fNIRS (functional near-infrared spectroscopy), energy,
prefrontal lobe
INTRODUCTION
Emotional labor is the act of regulating one’s emotion to conform to organizational standards.
Presently, it is central to numerous service occupations that employees are the first point of
contact that customers have with the organization. Scholars have been continually seeking to
understand the emotional labor process. Some have proposed that emotional labor consists of
three components (Grandey and Gabriel, 2015): emotional requirement, emotional regulation,
and emotional performance. The emotional requirements typically involve the “integrative” goal
of showing positive displays and hiding negative ones [although, for certain occupations (e.g.,
teachers), these emotional requirements may involve negative or neutral displays; Schutz and
Lee, 2014]. Emotional regulation refers to the effort expended by employees to comply with the
socioemotional demands of the job. Researchers have pointed out three essential strategies of
emotional regulation: surface acting, deep acting, and genuine emotional labor. Surface acting
involves faking the required emotions: that is, when people engage in surface acting, they do
not actually try to feel the emotions they wish to portray. They may put on “fake smiles” or
1 May 2019 | Volume 13 | Article 151
Lu et al.
other required emotional displays that do not reflect their true
feelings. By contrast, deep acting involves putting effort into
actually feeling and expressing the required emotions. When
engaged in deep acting, people attempt to modify feelings to
match the required display rules. Ashforth and Humphrey (1993)
argue that, in addition to surface acting and deep acting, that
there is a third form of emotional labor: spontaneous and genuine
emotional labor, wherein the person expresses naturally felt
emotions that align with the emotional display rules; in this
case, no acting is required (Diefendorff et al., 2005). Emotional
labor research has primarily focused on surface acting and deep
acting (Grandey et al., 2013; Gabriel and Diefendorff, 2015).
Finally, emotional performance means observable expressions of
emotional labor. Pugh (2001) found that emotional performance
was unrelated to employees’ reported moods, supporting the
idea that these displays were strategic rather than actual
expressions of feeling.
Since Hochschild (1983) coined the term “emotional labor,”
scholars have designed numerous studies to explore its
characteristics and mechanisms (e.g., Grandey, 2000). Among
them, the different strategies of emotional labor are one of the
most studied aspects—especially surface acting and deep acting
(Grandey et al., 2013; Xanthopoulou et al., 2018). Meta-analyses
have revealed that deep and surface acting have different effects
(Hülsheger and Schewe, 2011; Mesmer-Magnus et al., 2012),
and some interesting questions about these two strategies still
remain to be answered. For example, what are the cognitive
and social differences between these two strategies? Do they
consume identical amounts of psychological resources or does
one consume more? Most scholars believe that deep acting is
more psychologically taxing than surface acting is. However, until
now, there has been no real evidence to support this assertion.
The most common way of examining these two emotion
regulation strategies has been to ask employees to self-report the
extent to which they use each strategy, and then examine the
relationships between each strategy and various antecedents and
outcomes (Gabriel et al., 2015). However, self-report measures
often confound acting with affective motivational experiences;
this is true for both surface acting (i.e., confounded with negative
affect/stress) and deep acting (i.e., confounded with motivation).
Furthermore, self-reported results might simply be reflecting
effects of participants’ memory, and not emotional labor per se.
For this reason, some researchers have tried experimental
methods to explore differences between surface acting and
deep acting. For example, Shulei and Miner (2006) presented
participants with differing sets of instructions, before they
watched an emotion-eliciting film, to induce surface acting and
deep acting. The authors found that (1) the participants devoted
significant effort and found it difficult to perform both surface
and deep acting, but deep acting engaged more attention, and
(2) both surface acting and deep acting led to decreased sadness,
while deep acting led to a stronger physiological response (e.g.,
greater increase in heart rate). In another experimental study,
Gabriel and Diefendorff (2015) utilized a call center simulation to
examine how shifts in customer incivility influenced continuous
measures of participants’ felt emotions, surface acting, deep
acting, and vocal tone during a single interaction. They found
Frontiers in Human Neuroscience | www.frontiersin.org
Surface and Deep Acting Consumption
that surface acting and deep acting were used simultaneously to
manage emotional labor demands. Although these two studies
both used experimental methods, they still relied on participants’
self-ratings of surface acting and deep acting; in other words, they
could not measure the effort used for emotional labor directly.
Thus, we began to wonder if there is a way of measuring the effort
used for emotional labor directly. The answer may be “yes.”
According to the definition of emotional labor, which refers
to one kind of emotional regulation, so for surface acting and
deep acting, they can be seen as the two emotional regulation
strategies (Grandey, 2000; Grandey and Melloy, 2017). Gross
(1998) suggested that the emotional regulation strategies could
be categorized as antecedent-focused and response-focused, which
Grandey (2000) in turn connected to surface and deep acting;
deep acting is considered a form of reappraisal, which is linked to
antecedent-focused regulation, whereas surface acting is linked
to suppression, which is a type of response-focused regulation.
A large number of neuroimaging studies have been conducted
to examine the neural functions underlying the cognitive control
of emotion (Hare et al., 2005; Goldin et al., 2008; Etkin
et al., 2015; Zilverstand et al., 2017; Vandekerckhove et al.,
2018). Accordingly, neuroimaging technology could be useful for
exploring the neural mechanisms of emotional labor. However,
there are few neural studies on emotional labor—why? We
infer that the most important reason is the limitations such
studies impose on participants’ actions; participants would have
to utilize various expressive actions so that emotional labor can
be experimentally studied, but most neuroimaging equipment
requires participants to remain as immobile as possible. In
most cases, this would rule out the use of neuroimaging
techniques for emotional labor. However, functional near-
infrared spectroscopy (fNIRS) can be used with less bodily
constraints than other imaging modalities (Scholkmann et al.,
2012). In fact, some researchers have used fNIRS to explore
cortical areas that are associated with spontaneous facial affective
expressions, human emotions, somatosensory stimulation, and
other functions (Masataka et al., 2015; Naseer and Hong, 2015;
Sun et al., 2015; Hong et al., 2017). Therefore, it may be suitable
for use in emotional labor. Moreover, it is silent (i.e., it has no
operating sounds), making it somewhat more comfortable for
humans (Masataka et al., 2015).
Zapf (2002) discussed surface acting and deep acting from the
perspective of action theory. According to this theory, deep acting
would partly involve conscious processes at the intellectual level
of action regulation. In other words, the person actively tries to
influence his or her inner feelings to bring them in line with
the emotions required by the organization. By contrast, surface
acting is more likely to be triggered at the level of flexible action
patterns—this implies that it is partly a routine process, and
does not necessarily involve conscious processes. For example,
despite feeling nothing, a salesperson might automatically display
a smile. Based on Zapf ’s view of deep and surface acting,
we hypothesized that people will need greater psychological
resources when engaging in deep acting than when engaging
in surface acting.
In this study, the purpose was to verify whether people would
need more psychological resources when engaging in deep acting
2 May 2019 | Volume 13 | Article 151
Lu et al.
than when engaging in surface acting. To induce use of the
different strategies, we used emotional facial expressions. Because
emotional labor involves the display of positive and negative
emotions, we presented participants with positive or negative
facial expressions and asked them to make the opposite facial
expression. This also enabled us to explore whether there was a
difference in the effort for these strategies between positive and
negative emotional expressions.
MATERIALS AND METHODS
Ethics Statement
All research procedures were approved by the Ethical Committee
of Guizhou Normal University, Guiyang, Guizhou province,
China. All participants provided written informed consent after
fully understanding the study.
Participants
Twenty male and 20 female senior undergraduates were recruited
from a university in Hunan province, China. All participants
came from a teacher training specialty (e.g., education or
psychology) and were healthy and right-handed. None of
them had a history of psychiatric or neurological disorders.
Participants were randomized to the surface acting or deep
acting condition. In the deep acting condition, one male and one
female participant’s fNIRS data were not correctly obtained, so
the data were excluded. The average age of all participants was
21.5 ± 1.4 years.
Instrument
We used a 24-fiber (38-channel) fNIRS system (SHIMADZU
LABNIRS System, Kyoto, Japan) in this study. Semiconductor
lasers with wavelengths of 780 nm, 805 nm, and 830 nm are
employed as light sources; every emitter in the LABNIRS system
emits the three wavelength light, and as well as every detector
measures the absorbance for each of these wavelengths. fNIRS
measures the amount of relative change from the initial value
of oxygenated hemoglobin (OxyHb), deoxygenated hemoglobin
(deOxyHb), and total hemoglobin (totalHb), using the near-
infrared rays. The increase in OxyHb and the concomitant
decrease in deOxyHb reflects an increase in local arteriolar
vasodilatation, which increases local cerebral blood flow and
cerebral blood volume, a mechanism known as neurovascular
coupling. This produces a change in the amount of light
absorbed by this tissue, which can be measured by near
infrared spectroscopy systems (Ferrari and Quaresima, 2012).
These measurements are used as surrogates of brain activation
(Bulgarelli et al., 2018). The more activated, the more OxyHb
increases, the more energy will be consumed. The adequate
depth of NIR light penetration (almost one half of the source-
detector distance) can be achieved using a source-detector
distance around 3 cm (Dehghani et al., 2009). Compared with
other brain imaging devices, fNIRS is minimally restrictive and
can accurately record measurements even when individuals are
moving, making it useful for studying brain activity under more
“natural” conditions. It is also totally non-invasive.
Frontiers in Human Neuroscience | www.frontiersin.org
Surface and Deep Acting Consumption
FIGURE 1 | Functional near-infrared spectroscopy (fNIRS) settings. Ten
emitters and 10 detectors were deployed with reference to the International
10–20 system, with a total of 31 channels. For channel 10 and channel 13, the
distance of emitter and receptor was over 3 cm, so the measurement results
were canceled. The “cross” mark was used to represent the canceled results.
Numerous recent studies have identified the prefrontal cortex
(PFC) as a key region for the induction and regulation of
emotional responses (Doi et al., 2013). As our fNIRS instruments
could measure activity over a limited number of channels, we
decided to affix the optodes of the fNIRS primarily on the
PFC. Based on that, participants would make facial expression,
and refer to a study by Hong et al. (2015). Therefore, we also
placed some optodes on the pre-motor and supplementary motor
cortexes. The details are shown in Figure 1. The optodes of fNIRS
were fixed on a full head cap designed with reference to the
International 10–20 system. The sampling rate was set to 30 Hz.
We determined the anatomical locations of the optodes in
relation to the standard head landmarks, including the nasion,
Cz, left tragus (T3), and right tragus (T4), using a FASTTRAK 3D
tracking system (Polhemus). The Montreal Neurological Institute
(MNI) coordinates (Mazziotta et al., 2002) for the channels were
obtained using the NIRS-SPM software (Singh et al., 2005; Ye
et al., 2009) with MATLAB 2013b (Mathworks, Natick, MA,
United States). The detailed MNI locations are shown in Figure 2
and in Table 1.
Materials
We selected 20 positive, 20 neutral, and 20 negative emotional
face pictures from the Chinese Facial Affective Picture System
(CFAPS; Bai et al., 2005) using CFAPS normative data. Among
them, eight positive, eight neutral, and eight negative pictures
were chosen to be used in the exercise phase, while the remaining
were chosen to use in the formal phase. Emotion ratings were
based on a nine-point rating scale. The average emotional valence
rating of the selected pictures was 6.82 for the positive pictures,
4.32 for the neutral, and 2.34 for the negative. Half of all the
pictures were of men, and the others were of women.
Questionnaire
To identify whether participants had engaged with the tasks, we
asked participants to complete a 4-item questionnaire after the
3 May 2019 | Volume 13 | Article 151
Lu et al.
FIGURE 2 | MNI locations for the channels.
experiment, as follows: (1) What did you think about the difficulty
of the experiment? (2) How would rate the degree of effort put
into the experiment? (3) How would you rate your degree of
concentration on the experiment? (4) How would you rate your
performance in executing the experimental instructions? Each
item was rated on a six-point scale.
Experimental Procedure
Participants were instructed to sit on a chair in front of a 17-in
(32 × 24 cm) monitor. The distance between each participant
and the monitor was set to around 70 cm. The fNIRS equipment
was attached to participants’ heads. The experiment was divided
into the exercise and formal experimental phases. In the exercise
phase, participants practiced how to respond according to the
instructions until they completely understood it. There were 24
facial expression pictures (eight positive, eight neutral, and eight
negative), so in this phase, the participant would experience 24
trials under the set condition. Thereafter, the formal experiment
began. Before the formal experiment, participants were instructed
to avoid any head and body movements as much as possible
TABLE 1 | Registered positions of fNIRS measurement channels on the standard
brain MRI atlas.
Brodmann area Channel
BA 6 29, 30, 31
BA 8 24, 25, 26, 28
BA 9 15, 16, 17, 19, 20, 21, 22, 23, 27
BA 10 2, 3, 5, 6, 7, 8, 9, 11, 12
BA 11 1, 4
BA 45 18
BA 46 10, 13, 14
The channel numbers are defined in Figures 1, 2.
Frontiers in Human Neuroscience | www.frontiersin.org
Surface and Deep Acting Consumption
while the fNIRS was operating. Subsequently, emotional face
pictures were presented on the screen (315 × 356 pixels), and
participants were asked to make the opposite facial expression
of the pictures they viewed (see Figure 3 for details). In this
phase, the participants would experience 36 trials (12 positive, 12
neutral, and 12 negative). The facial expression of the participants
was video-monitored, in order to confirm that the correct
facial expressions were displayed. To minimize the distraction
of the participants, the experimental task was implemented in
a dimly lit room.
Surface Acting Condition
In this condition, we used the followed instructions to ensure that
participants performed surface acting:
Welcome to the experiment!
First you will see a “+” sign in the middle of the screen; please
look at it and keep your mind calm. After that, pictures of people
with positive, negative, and neutral emotional expressions will be
displayed on the screen. While focusing on the pictures, please
adopt the facial expression opposite to the one presented in order to
pretend to others that you are actually feeling that opposite emotion,
and keep that expression until the picture disappears. If a neutral
expression is displayed, please do not make any expression.
Deep Acting Condition
As with the surface acting condition, we used instructions to
induce participants to perform deep acting:
Welcome to the experiment!
First you will see a “+” sign in the middle of the screen; please
look at it and keep your mind calm. After that, pictures of people
with positive, negative, and neutral emotional expressions will be
displayed on the screen. While focusing on the pictures, please
adopt the facial expression opposite to the one presented, and try
to actually experience the emotion you are expressing on your face.
Please maintain that emotional state until the picture disappears. If
a neutral expression is displayed, please do not make any expression.
Data Preprocessing
As the distance between the emitting and detecting optodes
was over 3 cm, channel 10 was interpolated by channels
1, 5, 6, 14, 15, and 19; and channel 13 was interpolated
by channels 4, 8, 9, 17, 18, and 22. For this analysis, we
chose to focus on changes in the concentration of OxyHb,
as it is regarded as the most sensitive measure of changes
in regional cerebral blood flow (Hoshi et al., 2001). Using
the NIRS_SPM_V4_r1 toolbox (Ye et al., 2009), SPM8, and
MATLAB 2013b (Mathworks, Natick, MA, United States), we
processed participants’ individual data: (1) High-pass filtering:
we selected a discrete cosine transform (DCT)-based on
detrending algorithm. Based on DCT, the high-pass filtering was
implemented by a specified cut-off of 40 s (frequency = 0.025 Hz)
to remove unknown global trends due to breathing, cardiac
events, vasomotion, or other experimental errors. These trends
are the very low-frequency oscillations originating from slower
changes in systemic cardiovascular properties, e.g., blood
pressure (Gervain et al., 2011); (2) Low-pass filtering: according
4 May 2019 | Volume 13 | Article 151
Lu et al.
FIGURE 3 | One trial experimental procedure (written informed consent was obtained for the publication of this image).
to the NIRS-SPM developer’s recommendation (Ye et al.,
2009), we selected the precoloring method (Worsley and
Friston, 1995) and the hemodynamic response function (hrf )
to smooth the data. In this method, the intrinsic temporal
correlations are swamped by an imposed temporal correlation
structure, as smoothing the data with the temporal filtering will
attenuate high frequency components; hence, this is a “low-
pass filtering.” Since the transfer function of hrf is in the
frequencies of modeled neuronal signals, cut-off frequency did
not need to be set in this step; (3) Estimate: In the individual
analysis, general linear model (GLM) parameters and temporal
correlations were estimated, and beta values (average signal
value of each channel) obtained under four conditions were
calculated: condition 1 (a positive facial expression is presented);
condition 2 (a negative facial expression is presented); condition
3 (a neutral facial expression is presented); and condition 4
(no facial expression, just a “+” sign is presented). Using
these beta values, we conducted a group level analysis using
SPSS Statistics 20.
RESULTS
FNIRS Results
Activation of Channels
We compared the beta value of each channel with a 0 value. If the
p-value was significant, and the t-value of the one simple t-test
was positive, then the channel was defined to be activated; if the
t-value was negative, the channel was defined to be deactivated; if
the p-value was not significant, then the channel was defined to
not be activated.
To determine which channels were activated while
participants engaged in surface or deep acting, we subjected the
beta values estimated from the NIRS-SPM tool to a one sample
t-test and determined the level of activation when displaying
positive, negative, and neutral emotional face pictures. The
results are shown in Table 2. As shown in Table 2, according
to the BH_p results, when participants viewed a positive facial
expression (and therefore, made a negative expression), we
found that no channel was activated for surface acting. For
Frontiers in Human Neuroscience | www.frontiersin.org
Surface and Deep Acting Consumption
deep acting, channels 2, 9, 10, and 14 (BA10 and BA 46) were
activated. When participants viewed a negative facial expression
(and therefore, made a positive expression in response), those
in the surface acting group showed activations in channels 14,
29, and 30 (BA 6 and BA 46); by contrast, those in the deep
acting group showed no activated channels. When comparing
surface acting and deep acting, the number of activated
channels was slightly higher in the deep acting group than the
surface acting group.
Repeated Measures ANOVA
We then subjected the OxyHb data to a 2 × 3 repeated measures
analysis of variance (rmANOVA). The independent variables
were group and facial expressions. Group was a between-subject
variable, and included two levels: surface acting and deep acting.
Facial expression was a within-subject variable, and had three
levels: positive, negative, and neutral valence pictures. The results
indicated a non-significant interaction of the group and facial
expressions for all channels (all P > 0.05), and a non-significant
main effect of the group expression (all P > 0.05). However,
the main effect of the facial expression was significant for some
channels. We further merged the surface acting (n = 20) and
deep acting group (n = 18) into one group (n = 38), and
used facial expression as an independent variable to run a one-
way repeated measures ANOVA. The results showed that facial
expression was significant for some channels. For details, please
see Table 3. Specifically, according to BH_p results, a significant
effect of facial expression was found for channel 1 (located in
left orbitofrontal area, BA11); channels 5 (located in the left
frontopolar area, BA10); and channels 10, 14 [located in the
left dorsolateral PFC (L-DLFPC), BA46, please see Figure 2]. In
addition, in channels 24, 25, and 28 (located in the frontal eye
fields, BA8) and channels 29, 30, and 31 (located in the pre-motor
and supplementary motor cortexes, BA6), facial expression had a
significant effect.
At channels 5, 10, and 14, the OxyHb concentration change
for negative and positive facial expressions were significantly
higher than those for neutral facial expressions. Furthermore, at
channels 28, 29, 30, and 31, the OxyHb concentration change was
higher for positive facial expressions than for negative or neutral
5 May 2019 | Volume 13 | Article 151
Lu et al. Surface and Deep Acting Consumption

TABLE 2 | Comparison of activated channels among different conditions.

Surface acting (n = 20) Neutral (n = 38) Deep acting (n = 18)

Positive–negative Negative–positive Positive–negative Negative–positive

Channel t P BH_p t p BH_p t p t p BH_p t p BH_p

1 1.676 0.110 0.341 −0.398 0.695 0.798 −1.313 0.197 2.626 0.018∗ 0.110 1.521 0.147 0.301
2 −0.325 0.749 0.987 0.090 0.930 0.935 −0.719 0.477 3.348 0.004∗∗ 0.041∗ −0.398 0.696 0.768
5 1.431 0.169 0.475 0.786 0.441 0.624 −0.841 0.406 1.639 0.120 0.265 2.336 0.032∗ 0.138
9 −0.22 0.982 0.987 1.275 0.218 0.562 −0.225 0.823 3.330 0.004∗∗ 0.041∗ 0.951 0.355 0.440
10 1.910 0.071 0.276 0.882 0.389 0.624 −0.578 0.567 3.139 0.006∗∗ 0.046∗ 2.416 0.027∗ 0.138
14 2.327 0.031 0.265 3.245 0.004∗∗ 0.044∗ 1.048 0.301 4.440 0.000∗∗∗ 0.011∗ 3.571 0.002∗∗ 0.073
15 1.083 0.292 0.620 0.992 0.334 0.624 0.897 0.376 2.255 0.038∗ 0.167 2.916 0.010∗ 0.133
18 1.766 0.093 0.322 −0.427 0.674 0.798 1.043 0.304 2.436 0.026∗ 0.135 0.367 0.718 0.768
23 −0.856 0.403 0.734 2.218 0.047∗ 0.272 1.158 0.254 1.386 0.184 0.380 2.777 0.013∗ 0.133
24 −2.283 0.034∗ 0.265 1.334 0.198 0.562 −0.181 0.857 −1.769 0.095 0.263 1.237 0.233 0.344
25 −2.011 0.059 0.266 2.247 0.037∗ 0.272 −0.571 0.571 −0.319 0.754 0.899 1.355 0.193 0.340
26 −2.615 0.017∗ 0.264 0.973 0.343 0.624 −0.162 0.872 0.060 0.953 0.992 0.977 0.342 0.440
28 −2.811 0.011∗ 0.264 1.918 0.070 0.272 0.951 0.348 −1.318 0.205 0.397 2.282 0.036∗ 0.138
29 −1.999 0.060 0.266 3.519 0.002∗∗ 0.035∗ −0.659 0.514 −0.627 0.539 0.795 2.509 0.023∗ 0.138
30 −2.135 0.046∗ 0.266 4.306 0.000∗∗∗ 0.012∗ −0.683 0.499 0.010 0.992 0.992 2.061 0.055 0.155
31 −1.370 0.187 0.482 1.963 0.064 0.272 0.579 0.566 −0.541 0.596 0.839 2.626 0.018∗ 0.137

BH_p is the multiple comparison p-value correction using the Benjamini–Hochberg procedure; < 0.001;
∗∗∗ p < 0.01;
∗∗ p ∗p < 0.05. The asterisk marks annotation for
the following tables are the same as this table. Only the channels with significant p-value are displayed in this table.

TABLE 3 | Main effects of facial expression (n = 38).

Channel F p BH_p η2 PHMC Channel F p BH_p η2 PHMC

1 6.473 0.004∗∗ 0.018∗ 0.27 a>c 17 0.071 0.931 0.931 0.004 None
2 0.861 0.431 0.534 0.047 None 18 2.693 0.082 0.146 0.133 None
3 1.087 0.348 0.490 0.058 None 19 3.020 0.062 0.124 0.147 None
4 3.314 0.048∗ 0.106 0.159 None 20 3.420 0.044∗ 0.105 0.163 None
5 7.621 0.002∗∗ 0.012∗ 0.303 a > c, b > c 21 2.115 0.136 0.211 0.108 None
6 2.976 0.064 0.124 0.145 None 22 0.244 0.785 0.811 0.014 None
7 0.997 0.379 0.490 0.054 None 23 2.649 0.085 0.146 0.131 None
8 1.023 0.370 0.490 0.055 None 24 5.908 0.006∗∗ 0.021∗ 0.252 a < b, a < c
9 0.747 0.481 0.552 0.041 None 25 6.172 0.005∗∗ 0.019∗ 0.261 a < b, a < c
10 7.822 0.002∗∗ 0.012∗ 0.309 a > c, b > c 26 4.433 0.019∗ 0.054 0.202 None
11 3.922 0.029∗ 0.075 0.183 None 27 2.514 0.095 0.156 0.126 None
12 0.686 0.510 0.565 0.038 None 28 10.775 0.000∗∗∗ 0.002∗∗ 0.381 a < b, a < c, b > c
13 0.357 0.702 0.750 0.020 None 29 13.710 0.000∗∗∗ 0.000∗∗∗ 0.439 a < b, a < c, b > c
14 6.903 0.003∗∗ 0.015∗ 0.283 a > c, b > c 30 14.414 0.000∗∗∗ 0.000∗∗∗ 0.452 a < b, b > c
15 1.712 0.195 0.288 0.089 None 31 5.599 0.008∗∗ 0.025∗ 0.242 a < b, b > c
16 0.775 0.468 0.552 0.042 None

a, presented with a positive facial expression picture (requiring participants to make a negative expression); b, presented with a negative facial expression picture
(requiring participants to make a positive expression); c, presented with a neutral facial expression picture (requiring participant to not make any expression). PHMC,
post hoc multiple comparisons.

facial expressions. In channels 24, 25, 28, and 29, the OxyHb
concentration change was lower for negative facial expressions
than for positive and neutral facial expressions.
To further visualize the OxyHb concentration change
differences between positive, negative, and neutral expressions,
we used the results of the main effects analysis to calculate
t-values of the comparison between condition 1 and condition
3 and between condition 2 and condition 3. Then, we used the
Topoeasy toolbox (Tian et al., 2013) of MATLAB 2013b to plot a
topographical picture of the significant t-values (see Figure 4).
As shown in the topographical maps, compared to presenting
a neutral face picture (thereby requiring participants to make
no emotional expression), presenting positive and negative face
pictures significantly increased OxyHb concentration in the left
front and left middle areas of the PFC. Additionally, when
presenting a positive face picture (thereby requiring participants

Frontiers in Human Neuroscience | www.frontiersin.org 6 May 2019 | Volume 13 | Article 151

Lu et al.
FIGURE 4 | Topographical map of significant t-values (P < 0.05). Con 1, presenting a positive emotional face picture and asking participants to make a negative
expression; Con 2, presenting a negative emotional face picture and asking participants to make a positive expression; and Con 3, presenting a neutral face picture
and asking participants to not make any expressions.
to make a negative expression), the OxyHb concentration
significantly decreased in the left pre-motor and supplementary
motor cortex near the rear area of the PFC. Finally, when
presenting a negative face picture (thereby requiring participants
to make a positive expression), the OxyHb density significantly
increased in the pre-motor and supplementary motor cortex near
the rear area of PFC.
Questionnaire Results
Using the Student’s t-test, we examined participants’ subjective
feelings after completing the experiment, which could be a
supplement to the results of fNIRS. The results are shown
in Table 4.
For the means of all four questionnaire items, we found
no significant difference between the surface acting and
deep acting groups.
DISCUSSION
Energy Consumption of Surface and
Deep Acting
The main hypothesis of our study, which was based on action
theory, is that surface acting and deep acting consume different
amounts of psychological resources—particularly, deep acting
leads to greater energy consumption. However, the study results
appear somewhat contradictory. When we consider only the
number of activated areas in the brain, deep acting activated
more channels than surface acting did, thus supporting our study
hypothesis. However, when looking at the ANOVA results, we
observed neither significant differences in activation between
surface and deep acting nor a significant interaction of the group
and facial expression. In other words, these results do not support
our hypothesis. Thus, when using a more strict criterion (the
ANOVA results), it is possible that there is no difference in
energy consumption between surface and deep acting in the
Frontiers in Human Neuroscience | www.frontiersin.org
Surface and Deep Acting Consumption
PFC based on this study. However, this does not mean that
we can absolutely make a conclusion: there is no difference of
energy consumption between surface and deep acting. This is
because in addition to the PFC, other brain regions may also
be associated with emotional labor processes, such as subcortical
tissue amygdala, hippocampus, and so on. In addition, in this
study, we only used the instruction to ask participants to perform
surface acting and deep acting, and used a video monitor to
record their facial expressions. However, for the deep acting, we
did not monitor whether the participants were experiencing the
emotion corresponding to their facial expression (in fact it is very
difficult to do). This made it difficult to confirm whether the
participants really performed deep acting.
In addition to action theory, we might consider the results
in light of the conservation of resources theory. According to
this theory, surface acting, because it involves the suppression
of emotions, consumes more resources than deep acting does
(Richards and Gross, 2000), making it a strategy with greater
cognitive investment (Brotheridge and Lee, 2002). However, our
results do not appear to support the conservation of resources
theory. We suggest three possible reasons for this. First, there
might in fact be no significant difference in energy consumption
TABLE 4 | Status of participants after completing the experiment.
Comparison between groups n = 20 n = 18
Surface acting Deep acting
M ± SD M ± SD t
Q1 1.750 ± 0.910 1.944 ± 0.938 −0.648
Q2 1.800 ± 1.005 1.889 ± 1.023 −0.270
Q3 4.100 ± 0.788 4.167 ± 1.043 −0.224
Q4 5.150 ± 0.745 5.056 ± 0.725 0.395
Q1, Q2, Q3, and Q4 represent the four items on the questionnaire.
7 May 2019 | Volume 13 | Article 151
Lu et al.
between deep and surface acting (at least in the short term).
This is backed by the questionnaire results: the mean score for
the question “How would you rate the degree of effort put into
the experiment?” was slightly higher in the deep acting group
than in the surface acting group, but not to a significant degree
(see Table 1). In other words, based on participants’ subjective
feelings, there was no difference in deep acting and surface acting
in terms of effort. Second, the fNIRS can only measure the outer
cortex, surface of the brain. However, surface and deep acting
may also be involved with the deeper brain structure. For this,
a fNIRS study cannot discover whole brain activities related to
emotional labor, and it remains difficult to verify the action theory
and the conservation of resource theory. Third, emotional labor is
such a complex activity that it is possible that we have not actually
measured its energy consumption in this study. We simulated
only the surface and deep acting strategies of emotional labor; in
real emotional labor, the goals of the organization are exceedingly
important factors to consider, as they will doubtlessly influence
the emotional labor process (including surface and deep acting).
With our design, we would have found it difficult to present
participants with a definite organizational goal.
Differences Between Making Negative
and Positive Facial Expressions
We found that facial expression had a significant main effect for
some channels (see Table 3). To clarify the role of emotional
valence, we compared the changes in OxyHb concentration
between the presentation of a positive facial expression and
the presentation of a neutral facial expression, and between
the presentation of a negative facial expression and a neutral
facial expression. Compared with the neutral facial expression,
presenting either the positive or negative facial expression (and
making the corresponding opposite facial expression in response)
led to a notable activation in the left front and left middle parts of
the PFC (BA46 and BA10). According to the valence hypothesis,
the left PFC is activated during positive emotions, whereas the
right PFC is activated during negative emotions (Silberman
and Weingartner, 1986; Jansari et al., 2011). Using functional
magnetic resonance imaging (fMRI), Goldin et al. (2008) found
that the right ventrolateral PFC (VLPFC) was significantly
activated during emotional suppression while viewing disgusting
images. Kreplin and Fairclough (2013), in applying fNIRS during
the viewing of visual art, selected to induce positive and negative
emotions and found a significantly higher increase in OxyHb in
the medial rostral PFC (rPFC) when viewing positive images,
as compared to viewing negative images. Furthermore, Rodrigo
et al. (2016), also applying fNIRS, found that when viewing fearful
as compared to neutral faces, participants demonstrated higher
levels of activation within the right medial PFC. On the other
hand, participants demonstrated lower levels of activation within
the left medial PFC and left lateral PFC when viewing fearful
faces, as compared to neutral faces. Our results differ from both
these previous studies and do not support the valence hypothesis.
However, according to the affective workspace hypothesis, the
activity patterns of the same core neural network implement
both positive and negative valences, which are determined by
Frontiers in Human Neuroscience | www.frontiersin.org
Surface and Deep Acting Consumption
differences in the pattern of activation (Barrett and Bliss-Moreau,
2009; Trambaiolli et al., 2018). Lane et al. (1997a,b) presented
participants with happy, sad, and disgusting film clips, and asked
them to feel the relevant target emotion. They found increased
activity in the left medial PFC for both positive and negative
emotions, measured using PET. Honda et al. (2018) measured
the change in OxyHb concentration in the PFC while participants
commented on a disgusting film they had seen, and found
significant differences in OxyHb concentration in the left PFC
between when they engaged in emotional suppression and when
they did not. Our study results are in line with the studies of Lane
et al. (1997a,b) and Honda et al. (2018), and support the affective
workspace hypothesis.
Additionally, compared to making no facial expressions, when
participants engaged in a positive facial expression, the posterior
frontal lobe (mainly the frontal eye fields and supplementary
motor cortex) showed activity; however, this was not found
when they made a negative facial expression. This finding
might indicate that positive facial expressions are related to
stronger motion control. Lane et al. (1997a) also found that
happiness was distinguished from sadness by greater activity in
the vicinity of the ventral medial frontal cortex. However, we
cannot make any conclusions based such few studies, so further
research is necessary.
Limitations
First, in this study, we only presented participants with emotional
face pictures, and asked them to display the opposite facial
expression to the pictures they saw, using instructions to
manipulate whether they would engage in surface or deep acting.
Accordingly, our findings may not reflect real-life emotional
labor. Further studies should consider motivational factors in
order to improve the external validity of the study. Then, in this
study, we primarily focused on PFC activity. Although the PFC
is the most important zone for emotion regulation (Braunstein
et al., 2017), emotional responses to visual stimuli are regulated
through neural pathways in multiple brain regions, including the
amygdala and PFC, and emotional labor involves other parts of
the brain (e.g., the motor zone of the parietal lobe). In order
to explore whether there is a difference in energy consumption
between surface acting and deep acting, more areas of the
brain should be monitored in the future. Finally, we did not
monitor whether the participants were experiencing the emotion
corresponding to their facial expression. This made it difficult
to confirm whether the participants really performed a deep
acting. So, in future studies, it is worth exploring how to use
behavioral and physiological indicators to measure participants’
surface acting and deep acting.
CONCLUSION
Based on the results of this study, we infer that deep acting and
surface acting may not show a significant difference in energy
consumption. Furthermore, engaging in positive and negative
emotional labor may rely on some of the same psychological
mechanisms, though there could be differences as well.
8 May 2019 | Volume 13 | Article 151
Lu et al.
ETHICS STATEMENT
All research procedures were approved by the Research Ethical
Committee of Guizhou Normal University Educational School
according to the Declaration of Helsinki. All participants
were given written informed consent after they fully
understand the study.
AUTHOR CONTRIBUTIONS
WW designed the experiments. YL, HZ, DL, and
DP collected the data. YL and GM analyzed the
REFERENCES
Ashforth, B. E., and Humphrey, R. H. (1993). Emotional labor in service roles: the
influence of identity. Acad. Manag. Rev. 18, 88–115. doi: 10.2307/258824
Bai, L., Ma, H., Huang, Y., and Luo, Y. (2005). The development of native chinese
affective picture system–A pretest in 46 college students. Chin. Mental Health J.
19, 719–722.
Barrett, L., and Bliss-Moreau, E. (2009). Affect as a psychological primitive. Adv.
Exp. Soc. Psychol. 41, 167–218. doi: 10.1016/S0065-2601(08)00404-8
Braunstein, L. M., Gross, J. J., and Ochsner, K. N. (2017). Explicit and implicit
emotion regulation: a multi-level framework. Soc. Cogn. Affect. Neurosci. 12,
1545–1557. doi: 10.1093/scan/nsx096
Brotheridge, C. M., and Lee, R. T. (2002). Testing a conservation of resources
model of the dynamics of emotional labor. J. Occup. Health Psychol. 7, 57–67.
doi: 10.1037//1076-8998.7.1.57
Bulgarelli, C., Blasi, A., Arridge, S., Powell, S., Klerk, C. D., Southgate, V., et al.
(2018). Dynamic causal modelling on infant fNIRS data: a validation study
on a simultaneously recorded fNIRS-fMRI dataset. Neuroimage 175, 413–424.
doi: 10.1016/j.neuroimage.2018.04.022
Dehghani, H., White, B. R., Zeff, B. W., Tizzard, A., and Culver, J. P. (2009).
Depth sensitivity and image reconstruction analysis of dense imaging arrays
for mapping brain function with diffuse optical tomography. Appl. Opt. 48,
D137–D143. doi: 10.1364/AO.48.00D137
Diefendorff, J. M., Croyle, M. H., and Gosserand, R. H. (2005). The dimensionality
and antecedents of emotional labor strategies. J. Vocat. Behav. 66, 339–357.
doi: 10.1016/j.jvb.2004.02.001
Doi, H., Nishitani, S., and Shinohara, K. (2013). NIRS as a tool for assaying
emotional function in the prefrontal cortex. Front. Hum. Neurosci. 7:770. doi:
10.3389/fnhum.2013.00770
Etkin, A., Büchel, C., and Gross, J. J. (2015). The neural bases of emotion regulation.
Nat. Rev. Neurosci. 16, 693–700. doi: 10.1038/nrn4044
Ferrari, M., and Quaresima, V. (2012). A brief review on the history of
human functional near-infrared spectroscopy (fNIRS) development and fields
of application. Neuroimage 63, 921–935. doi: 10.1016/j.neuroimage.2012.
03.049
Gabriel, A., and Diefendorff, J. (2015). Emotional labor dynamics: a momentary
approach. Acad. Manag. J. 58, 1804–1825. doi: 10.5465/amj.2013.1135
Gabriel, A. S., Daniels, M. A., Diefendorff, J. M., and Greguras, G. J. (2015).
Emotional labor actors: a latent profile analysis of emotional labor strategies.
J. Appl. Psychol. 100, 863–879. doi: 10.1037/a0037408
Gervain, J., Mehler, J., Werker, J. F., Nelson, C. A., Csibra, G., Lloyd-Fox, S.,
et al. (2011). Near-infrared spectroscopy: a report from the McDonnell infant
methodology consortium. Dev. Cogn. Neurosci. 1, 22–46. doi: 10.1016/j.dcn.
2010.07.004
Goldin, P. R., Mcrae, K., Ramel, W., and Gross, J. J. (2008). The neural bases
of emotion regulation: reappraisal and suppression of negative emotion. Biol.
Psychiatry 63, 577–586. doi: 10.1016/j.biopsych.2007.05.031
Grandey, A. A. (2000). Emotional regulation in the workplace: a new way to
conceptualize emotional labor. J. Occup. Health Psychol. 5, 95–110. doi: 10.1037/
/1076-8998.5.1.95
Frontiers in Human Neuroscience | www.frontiersin.org
Surface and Deep Acting Consumption
data. YL, WW, and SZ wrote the main manuscript.
YL and WW prepared the figures. All authors
reviewed the manuscript.
FUNDING
This study was supported by Startup Foundation for Advanced
Talents in Guizhou Normal University for WW; and also
funded by Philosophy and Social Sciences Key Research
Base Open Fund Project of Hunan Provincial University
(13K093) and by Hunan Provincial Social Science Fund
Project (13YBA138).
Grandey, A. A., Diefendorff, J. M., and Rupp, D. E. (2013). “Bringing emotional
labor into focus: A review and integration of three research lenses,” in
Organization and Management. Emotional Labor in the 21st Century: Diverse
Perspectives on Emotion Regulation at Work, eds A. A. Grandey, J. M.
Diefendorff, and D. E. Rupp (New York, NY: Routledge/Taylor & Francis
Group), 3–27.
Grandey, A. A., and Gabriel, A. S. (2015). Emotional labor at a crossroads: where
do we go from here? Ann. Rev. Organ. Psychol. Organ. Behav. 2, 323–349.
doi: 10.1146/annurev-orgpsych-032414-111400
Grandey, A. A., and Melloy, R. C. (2017). The State of the heart: emotional labor as
emotion regulation reviewed and revised. J. Occup. Health Psychol. 22, 407–422.
doi: 10.1037/ocp0000067
Gross, J. J. (1998). The emerging field of emotion regulation: An integrative review.
Rev. Gen. Psychol. 2, 271–299. doi: 10.1037/1089-2680.2.3.271
Hare, T. A., Tottenham, N., Davidson, M. C., Glover, G. H., and Casey, B. J. (2005).
Contributions of amygdala and striatal activity in emotion regulation. Biol.
Psychiatry 57, 624–632. doi: 10.1016/j.biopsych.2004.12.038
Hochschild, A. R. (1983). The Managed Heart: Commercialization of Human
Feeling. Berkeley, CA: University of California Press.
Honda, M., Tanaka, H., Sakti, S., and Nakamura, S. (2018). “Detecting suppression
of negative emotion by time series change of cerebral blood flow using fNIRS,”
in Proceedings of the 2018 IEEE EMBS International Conference on Biomedical
& Health Informatics (BHI), (Las Vegas, NV: IEEE).
Hong, K. S., Bhutta, M. R., Liu, X., and Shin, Y. I. (2017). Classification of
somatosensory cortex activities using fNIRS. Behav. Brain Res. 333, 225–234.
doi: 10.1016/j.bbr.2017.06.034
Hong, K. S., Naseer, N., and Kim, Y. H. (2015). Classification of prefrontal and
motor cortex signals for three-class fNIRS–BCI. Neurosci. Lett. 587, 87–92.
doi: 10.1016/j.neulet.2014.12.029
Hoshi, Y., Kobayashi, N., and Tamura, M. (2001). Interpretation of near-
infrared spectroscopy signals: a study with a newly developed perfused rat
brain model. J. Appl. Physiol. 40, 1657–1662. doi: 10.1152/jappl.2001.90.5.
1657
Hülsheger, U. R., and Schewe, A. F. (2011). On the costs and benefits of emotional
labor: a meta-analysis of three decades of research. J. Occup. Health Psychol. 16,
361–389. doi: 10.1037/a0022876
Jansari, A., Rodway, P., and Goncalves, S. (2011). Identifying facial emotions:
valence specific effects and an exploration of the effects of viewer gender. Brain
Cogn. 76, 415–423. doi: 10.1016/j.bandc.2011.03.009
Kreplin, U., and Fairclough, S. H. (2013). Activation of the rostromedial prefrontal
cortex during the experience of positive emotion in the context of esthetic
experience. An fNIRS study. Front. Hum. Neurosci. 7:879. doi: 10.3389/fnhum.
2013.00879
Lane, R. D., Reiman, E. M., Ahern, G. L., Schwartz, G. E., and Davidson, R. J.
(1997a). Neuroanatomical correlates of happiness, sadness, and disgust. Am.
J. Psychiatry 154, 926–933. doi: 10.1176/ajp.154.7.926
Lane, R. D., Reiman, E. M., Bradley, M. M., Lang, P. J., Ahern, G. L., Davidson,
R. J., et al. (1997b). Neuroanatomical correlates of pleasant and unpleasant
emotion. Neuropsychologia 35, 1437–1444. doi: 10.1016/S0028-3932(97)
00070-75
9 May 2019 | Volume 13 | Article 151
Lu et al.
Masataka, N., Perlovsky, L., and Hiraki, K. (2015). Near-infrared spectroscopy
(NIRS) in functional research of prefrontal cortex. Front. Hum. Neurosci. 9:274.
doi: 10.3389/fnhum.2015.00274
Mazziotta, J., Toga, A., Evans, A., Fox, P., Lancaster, J., Zilles, K., et al. (2002).
A probabilistic atlas and reference system for the human brain: international
consortium for brain mapping (ICBM). Philos. Trans. R Soc. Lond. B Biol. Sci.
356, 1293–1322. doi: 10.1098/rstb.2001.0915
Mesmer-Magnus, J. R., DeChurch, L. A., and Wax, A. (2012). Moving emotional
labor beyond surface and deep acting a discordance–congruence perspective.
Organ. Psychol. Rev. 2, 6–53. doi: 10.1177/2041386611417746
Naseer, N., and Hong, K. S. (2015). Decoding answers to four-choice questions
using functional near infrared spectroscopy. J. Near Infrar. Spectrosc. 23, 23–31.
doi: 10.1255/jnirs.1145
Pugh, S. D. (2001). Service with a smile: emotional contagion in the service
encounter. Acad. Manag. J. 44, 1018–1027. doi: 10.2307/3069445
Richards, J. M., and Gross, J. J. (2000). Emotion regulation and memory: the
cognitive costs of keeping one’s cool. J. Pers. Soc. Psychol. 79, 410–424.
doi: 10.1037/0022-3514.79.3.410
Rodrigo, A. H., Ayaz, H., and Ruocco, A. C. (2016). “Examining the Neural
Correlates of Incidental Facial Emotion Encoding within the Prefrontal
Cortex Using Functional Near-Infrared Spectroscopy,” in Foundations of
Augmented Cognition: Neuroergonomics and Operational Neuroscience, eds
D. D. Schmorrow and C. M. Fidopiastis (Cham: Springer).
Scholkmann, F., Gerber, U., Wolf, M., and Wolf, U. (2012). End-tidal CO2: an
important parameter for a correct interpretation in functional brain studies
using speech tasks. Neuroimage 66, 71–79. doi: 10.1016/j.neuroimage.2012.
10.025
Schutz, P. A., and Lee, M. (2014). “Teacher emotion, emotional labor and
teacher identity,” in English as a Foreign Language Teacher Education: Current
Perspectives and Challenges, Vol. 27, eds J. de Dios and M. Agudo (Amsterdam:
Rodopi), 169–186.
Shulei, M., and Miner, H. (2006). Emotional labor; surface acting and deep acting,
which one is better? Acta Psychol. Sin. 38, 262–270.
Silberman, E. K., and Weingartner, H. (1986). Hemispheric lateralization of
functions related to emotion. Brain Cogn. 5, 322–353. doi: 10.1016/0278-
2626(86)90035-7
Singh, A. K., Okamoto, M., Dan, H., Jurcak, V., and Dan, I. (2005). Spatial
registration of multichannel multi-subject fNIRS data to MNI space without
MRI. Neuroimage 27, 842–851. doi: 10.1016/j.neuroimage.2005.05.019
Frontiers in Human Neuroscience | www.frontiersin.org
Surface and Deep Acting Consumption
Sun, Y., Ayaz, H., and Akansu, A. N. (2015). “Neural correlates of affective context
in facial expression analysis: A simultaneous EEG-fNIRS study,” in Proceedings
of the 2015 IEEE Global Conference on Signal and Information Processing
(GlobalSIP), (Orlando, FL: IEEE).
Tian, F., Lin, Z. J., and Liu, H. (2013). EasyTopo: A toolbox for rapid diffuse optical
topography based on a standard template of brain atlas. Proc. SPIE 8578:85782J.
doi: 10.1117/12.2003907
Trambaiolli, L. R., Biazoli, C. E., Cravo, A. M., and Sato, J. R. (2018). Predicting
affective valence using cortical hemodynamic signals. Sci. Rep. 8:5406. doi:
10.1038/s41598-018-23747-y
Vandekerckhove, M., Van Hecke, W., Quirin, M., and De Mey, J. (2018). Neural
pathways in ‘emotional approach’as experiential emotion regulation strategy.
Behav. Brain Res. doi: 10.1016/j.bbr.2018.01.001 [Epub ahead of print].
Worsley, K. J., and Friston, K. J. (1995). Analysis of fMRI time-series revisited–
again. Neuroimage 2, 173–181. doi: 10.1006/nimg.1995.1023
Xanthopoulou, D., Bakker, A. B., Oerlemans, W. G., and Koszucka, M. (2018).
Need for recovery after emotional labor: differential effects of daily deep and
surface acting. J. Organ. Behav. 39, 481–494. doi: 10.1002/job.2245
Ye, J. C., Tak, S., Jang, K. E., Jung, J., and Jang, J. (2009). NIRS-SPM: statistical
parametric mapping for near-infrared spectroscopy. Neuroimage 44, 428–447.
doi: 10.1016/j.neuroimage.2008.08.036
Zapf, D. (2002). Emotion work and psychological well-being: a review of the
literature and some conceptual considerations. Hum. Res. Manag. Rev. 12,
237–268. doi: 10.1016/s1053-4822(02)00048-7
Zilverstand, A., Parvaz, M. A., and Goldstein, R. Z. (2017). Neuroimaging cognitive
reappraisal in clinical populations to define neural targets for enhancing
emotion regulation. a systematic review. Neuroimage 151, 105–116. doi: 10.
1016/j.neuroimage.2016.06.009
Conflict of Interest Statement: The authors declare that the research was
conducted in the absence of any commercial or financial relationships that could
be construed as a potential conflict of interest.
Copyright © 2019 Lu, Wu, Mei, Zhao, Zhou, Li and Pan. This is an open-access
article distributed under the terms of the Creative Commons Attribution License
(CC BY). The use, distribution or reproduction in other forums is permitted, provided
the original author(s) and the copyright owner(s) are credited and that the original
publication in this journal is cited, in accordance with accepted academic practice. No
use, distribution or reproduction is permitted which does not comply with these terms.
10 May 2019 | Volume 13 | Article 151